266 Ann Ist Super Sanità 2010 | Vol. 46, No. 3: 266-273 DOI: 10.4415/ANN_10_03_07

Quantitative real-time PCR of enteric on c ern viruses in influent and effluent samples C from wastewater treatment plants in Italy e a lt h

H Giuseppina La Rosa, Manoochehr Pourshaban, Marcello Iaconelli and Michele Muscillo of

Dipartimento di Ambiente e Connessa Prevenzione Primaria, Istituto Superiore di Sanità, Rome, Italy ssues

I Summary. The prevalence of enteric viruses in wastewater, the efficacy of wastewater treatments in elimi- nating such viruses, and potential health risks from their release into the environment or by recycling of treated wastewaters, are very important issues in environmental microbiology. In this study we performed a quantitative TaqMan real-time PCR (polymerase chain reaction) analysis of enteric viruses on samples of influents and effluents from 5 wastewater treatment plants in and around Rome. Three epidemiologi-

a l vi ronment cally important, waterborne enteric viruses were analyzed: adenoviruses, enteroviruses and noroviruses n (GI and GII) and compared to classical bacterial indicators of fecal contamination. The concentration E of adenoviruses was the highest, in both raw and treated waters. Mean values in influents were ranked as follows: adenovirus > norovirus GI > norovirus GII > enterovirus. In effluents, the ranking was: adeno- virus > norovirus GI > enterovirus > norovirus GII. Removal efficiencies ranged from 35% (enterovirus) to 78% (norovirus GI), while removal efficiency for bacterial indicators was up to 99%. Since molecular quantification does not necessarily indicate an actual threat to human health, we proceeded to evaluate the infectivity of enterovirus particles in treated effluents through integrated cell culture and real-time PCR. Infectivity assays detected live virions in treated water, pointing to potential public health risks through the release of these viruses into the environment. A better understanding of viral presence and resistance to sewage purification processes have the potential of contributing to the effective management of risks linked to the recycling of treated wastewater, and its discharge into the environment. Key words: norovirus, adenovirus, enterovirus, sewage, real-time quantitative PCR assays.

Riassunto (Determinazione quantitativa mediante real-time PCR di virus enterici in campioni di acque reflue affluenti ed effluenti da impianti di trattamento dei liquami in Italia). La prevalenza dei virus enterici nei liquami urbani, la loro rimozione durante i trattamenti di depurazione delle acque reflue, e rischi po- tenziali per la salute legati alla contaminazione degli ambienti idrici o al riutilizzo delle acque depurate, sono argomenti di grande interesse in microbiologia ambientale. In questo studio abbiamo stimato la concentrazione di virus enterici, in campioni di reflui grezzi e depurati di 5 impianti di trattamento di reflui urbani localizzati in Roma e dintorni, utilizzando metodologie quantitative molecolari (TaqMan real-time PCR). Sono stati analizzati tre gruppi di virus enterici responsabili di epidemie di origine idrica (adenovirus, enterovirus e norovirus GI/GII), parallelamente ai classici indicatori batterici di contami- nazione fecale. Le concentrazioni virali più elevate sono state rilevate per gli adenovirus, sia nei reflui non depurati sia in quelli trattati. Nei reflui grezzi sono state osservate le seguenti concentrazioni medie: adenovirus > norovirus GI > norovirus GII > enterovirus; negli effluenti depurati: adenovirus > noro- virus GI > enterovirus > norovirus GII. L’efficienza di rimozione virale variava dal 35% (enterovirus) al 78% (norovirus GI), mentre per gli indicatori batterici raggiungeva il 99%. Poiché i metodi molecolari non forniscono indicazioni circa l’infettività dei virus rilevati, si è proceduto a valutare la capacità di replicazione delle particelle virali (limitatamente agli enterovirus) nei liquami depurati utilizzando un sistema integrato colture cellulari/real-time PCR. L’analisi ha rilevato particelle virali infettive, indicando potenziali rischi per la salute pubblica legati alla contaminazione dei corpi idrici recettori. Una migliore comprensione della resistenza dei virus ai processi di depurazione è importante per la gestione dei rischi legati al riutilizzo delle acque reflue trattate, o alla loro immissione nell’ambiente. Parole chiave: norovirus, adenovirus, enterovirus, liquami, PCR quantitativa real-time.

INTRODUCTION Most treated wastewater, as well as untreated sewage, Enteric viruses are largely excreted through fecal flow into the water environment and have the poten- matter, and current methods of sewage treatment tial to impact agricultural, recreational and drinking do not always effectively remove these organisms [1]. waters. Currently, bacterial indicators such as fecal

Address for correspondence: Giuseppina La Rosa, Dipartimento di Ambiente e Connessa Prevenzione Primaria, Istituto Superiore di Sanità, Viale Regina Elena 299, 00161 Rome, Italy. E-mail: [email protected]. Quantification of enteric viruses at WTPs 267

coliforms and enterococci are commonly used to method combining the sensitive conventional quan- assess water quality. These bacteria are simple and titative PCR with an infectivity assay. inexpensive to monitor, but have been shown to be less than ideal as indicators of fecal pollution. In on c ern fact, enteric human-pathogenic viruses are generally METHODS C more resistant than enteric bacteria to current meth- Sample collection ods of wastewater treatment [2, 3]. Different studies Five different municipal WTPs were chosen, all e a lt h having found that, despite treatment, enteric viruses situated in central Italy, within 30 km of each oth- H of

persist in high levels in wastewaters [4-7], proposed er. The WTPs in question are conventional activat- the use of a viral indicator of wastewater contami- ed sludge plants which receive waters from urban nation. Specifically, enteroviruses (EVs) and adeno- areas. Wastewater treatment typically includes grid ssues viruses (AdVs) have been suggested as indicators for separation, primary sedimentation, secondary bio- I monitoring the human fecal contamination of wa- logical treatment and final disinfection with chlo- ter and for determining the efficacy of disinfection rine, before effluents are discharged into the water treatments [8]. The monitoring of wastewater treat- environment. Daily flows in these plants range from ment plants (WTP) may prove a suitable approach 4800 to 750 000 cubic meters, and design capacity

for the study of circulating viruses in their respective from 60 000 to 1 100 000 population equivalents. a l vi ronment service areas [9], and the persistence of such viruses No industrial wastewater is treated in any of these n in treated effluents. facilities. E The objective of the present study was to assess, A total of 50 grab samples were collected (25 in- in both raw and treated sewages, the concentration flows and 25 outflows, 50 mL each). Sampling was of three epidemiologically important, waterborne carried out monthly, from May to September 2007. enteric viruses, AdVs, EVs and noroviruses (NoV Samples were collected in the morning on sunny GI and NoV GII), to gain insight into the infectious days, stored in sterile 50 mL Falcon tubes kept in potential of viral contamination of treated sewage, thermal bags at 4 °C, and delivered to the labora- and to obtain a rough estimate the efficacy of dis- tory, where they were processed within 24 hours. infection treatments. For the sake of reference, we also tested and quantified two enteric bacteria com- Concentration of viruses and RNA extraction monly used as fecal indicators, Escherichia coli and Viruses were concentrated as previously described enterococci. [16]. Briefly, samples were centrifuged at 3000 ×g at 4 EVs are single-stranded RNA, non-enveloped vi- °C for 10’. Supernatants were centrifuged at 200 000 ruses. Symptoms of EV infection include mild up- × g for 2 h. The pellets were resuspended in 1 mL of per respiratory illness, febrile rash, aseptic meningi- phosphate-buffered saline and the suspension aliq- tis, pleurodynia, encephalitis, acute flaccid paralysis, uoted and stored at -80 °C until genome extraction. paralytic poliomyelitis, gastroenteritis, myocarditis, To assess virus recovery, we spiked 50 mL of dis- pericarditis and diabetes [10, 11]. tilled water with 1.91 × 108 genome copies (GC) of Human AdVs are double-stranded DNA, non-en- Echovirus 2 (Cornelis strain, american type culture veloped viruses, with 52 serotypes grouped into six collection, ATCC, VR-32) and computed the ratio of species, from A to F. Symptoms of AdV infection GCs after/before concentration by real-time PCR. include gastroenteritis, pharyngitis, pneumonia, RNA and DNA were isolated from 100 µl of viral conjunctivitis, and meningoencephalitis [12]. pellet using the NucliSens miniMAG (BioMérieux NoVs, the most commonly identified cause of Italia, Florence, Italy) nucleic acid isolation kit, acute nonbacterial gastroenteritis in all age groups, based on the Boom nucleic acid extraction method contain a single-stranded, positive-sense RNA ge- [17] in combination with magnetic silica beads. A nome [13, 14]. These viruses are members of the NucliSens miniMag extractor (BioMérieux Italia, Caliciviridae family, and are classified into five ge- Italy) was used to collect and wash the magnetic nogroups, GI-GV [15]. silica particles, and to recover nucleic acids from the Each group of viruses was quantified using an ap- magnetic beads. Viral genomes were eluted from the propriate TaqMan real-time PCR (or RT-PCR) as- silica in 100 µl elution buffer and stored in aliquots say. Real-time PCR has been demonstrated to be a at -80 °C until use. powerful tool for nucleic acid quantification in water environments, thanks not only to its sensitivity and Bacteriological assays specificity, but also to the fact that, relative to other Upon arrival, samples were aliquoted and analyzed. quantitative tools, it is quick and easy to use. For E. coli, microtiter plates (BioRAD, Milan, Italy) The quantitative molecular test is unable to dis- were used, following the manufacturer’s protocol criminate between viable and inactivated virions, so “Miniaturized method for the enumeration of E. coli that the presence of a large quantity of viral par- in surface and waste waters MUG/EC”. Similarly, ticles does not necessarily indicate an actual threat for enterococci, microtiter plates were used following to human health. Therefore, we evaluated the infec- the manufacturer’s protocol “Miniaturized method tivity of EVs in treated effluents through integrated for the enumeration of enterococci in surface and cell culture and real-time PCR (ICC-RT-PCR), a waste waters MUD/SF”. Both these protocols are in 268 Giuseppina La Rosa, Manoochehr Pourshaban, Marcello Iaconelli, et al.

accordance with ISO procedures (ISO 9308-3, 1998, Biosystems, Monza, Italy) on a duplicate set using 10 ISO-7899-1, 1998). μl of the extracted genome. Reactions were carried out In both cases, bacterial counts, corresponding to in disposable optical 96-well PCR plates with flat tops on c ern the number of fluorescent wells, were expressed in and printed grids (Labcon, Petaluma, CA, USA) in a 25 C Most Probable Number (MPN) (Table 1), in accord- μl mixture using a SensiMix DNA kit for AdVs, and a ance with ISO 8199. SensiMix One-Step RNA kit (Quantace, London, UK) e a lt h for EVs and NoVs.

H Quantification standard For AdVs, the reaction mixture was heated at 50 °C of

We performed an absolute quantization to deter- for 2 min and then at 95 °C for 10 min; activation was mine the number of viral GCs contained in the sam- followed by a 40-cycle, two step process, each cycle ples, using an external calibration curve generated by consisting of denaturation at 95 °C for 15 s and an- ssues 8 1 I 10-fold serial dilutions (10 -10 ) of a standard con- nealing or extension at 60 °C for 1 min. For EVs and sisting of a linearized plasmid DNA for AdVs, and NoVs, the reaction mixture was initially incubated at of synthetic ssRNA strands for EVs and NoVs GI 42 °C for 45 min to transcribe RNA. Because of the and GII. The standards were diluted in nuclease-free extreme sensitivity of the quantitative PCR, we paid water containing carrier yeast tRNA (100 ng/μl). particular attention to the prevention of contamina-

a l vi ronment The quantification standards for NoV GI and GII, tion: separate rooms were used for the preparation or n previously constructed by our team, were 762- and mixing of reagents, sample processing, and gel electro- E 591-nucleotide ssRNA strands, respectively, con- phoresis; reagents and samples were stored in separate taining a partial sequence coding for RNA-depend- rooms; and the equipment used in each room was not ent RNA polymerase [18]. Quantitative analysis for used in other areas. Negative PCR mixture controls, NoVs was performed on the highly conserved re- and extraction controls were systematically used. gion between ORF1 and ORF2, using genogroup- Curves were generated by the Sequence Detection specific assays, able to amplify 96- and 98-bp frag- System software, plotting cycle threshold (Ct) ments, for genogroup I and II respectively, following against the number of viral GCs. Run acceptability a slightly modified version [18] of the TaqMan assay was defined as a correlation coefficient (R2) > 0.98 described by Jothikumar and collaborators [19]. and a slope between -3.6 and -3.1, corresponding to The calibration curve used for AdV quantization was reaction efficiencies between 90 and 110%, accord- a HindIII linearized plasmid prepared from a recom- ing to the equation: Efficiency = 10(-1/slope) –1 [24]. binant pCR4TOPO vector containing part of the hexon To verify the suitability of environmental samples gene [20]. TaqMan real-time PCR was performed using for amplification (absence of inhibitors), negative the assay described by Hernroth and collaborators [21], samples were retested, with a calibrated stock of with a minor modification [20], which amplify a 69-bp RNA or DNA standard (see above) added to the fragment of the conserved region of the hexon gene. real-time mixture. The standard RNA for EV quantification was pre- pared by our team as follows: a BamHI220-670 frag- Statistical analysis ment (451-bp) from a pVR106 plasmid [22] contain- Reproducibility of real time PCR was estimated ing the whole Poliovirus 1 genome was cloned into by assessing the intra- and interassay variability (i.e. the pGEM-4Z vector. The recombinant plasmid was between identical experiments performed in dupli- purified using Qiagen minipreps (Qiagen, Hilden, cate on the same day, and experiments performed on Germany). Having obtained the DNA standard, we different days). Percent coefficient of variation (% then proceeded to construct the RNA standard using CV) was used to quantify the variation of Ct values RiboMAX large scale RNA production systems-T7 across replicates of standard curve dilutions (Reed (Promega Italia, Milan, Italy) following the manu- et al. 2002). facturer’s instructions. A 513-nucleotide ssRNA was Quantification data were exported into a Microsoft T7-transcribed and quantified using the ND-1000 in- Excel file for subsequent statistical analysis. strument (Nanodrop, Wilmington, DE, USA). The Percent removal was calculated as follows: R (%) = concentrations of ssRNA were then converted into (MPCout – MPCin / MPCin) × 100, where MPCout copy number using the following equation: [GC/mL stands for mean pathogen concentration in effluent = C/MW * 6.02 × 1023], where C is RNA concentra- samples, and MPCin stands for mean pathogen tion (C = g/mL), and MW is the molecular weight. concentration in influent samples. Mean is intended Quantitative PCR was performed according to the as arithmetic mean for bacteria and mean trimmed assay described by Donaldson and collaborators [23], 20% for viruses. Additionally, maximum and mini- an assay based on TaqMan probe hydrolysis, ampli- mum values were calculated. fying 192 bp of the 5’ untranscribed region. Cell culture analysis and sequencing for human Table 2 shows the list of primers and PCRs used enteroviruses in this study. The infectivity of EVs in treated water samples was analyzed using human rhabdomyosarcoma (RD) Real-time PCR cells. After concentration and ultracentrifugation, Real-time PCRs were carried out in an ABI- 200 μL of viral suspension were purified by chloro- PRISM 7000 sequence detection instrument (Applied form extraction, and inoculated into a 25 cm2 tissue Quantification of enteric viruses at WTPs 269

culture flask. The inoculum was discarded and cells counts was observed in effluent waters as compared were overlaid with 4 mL of maintenance medium to influents. The mean removal of viable bacterial (Dulbecco’s Modified Eagle Medium, DMEM). particles in treated sewage was 98.9% and 98.5% for Cell cultures were incubated for 5 days in 5% CO2 E. coli and enterococci, respectively. on c ern at 37 °C and observed daily by optical microscope. C An echovirus 2 strain (VR-32, ATCC, USA) was Real-time PCR results used as a positive control; uninfected cells served as Virus concentrations in sewage samples were deter- e a lt h negative controls. The incubation and cell culture mined by quantitative real-time PCR, as described H of

facilities were physically separated from the PCR above. Slope values ranged from -3.2 to -3.5 indicating facility. After 5 days, cell culture flasks were scored a high PCR efficiency in these experiments (> 90%). for the presence of CPE and then frozen and thawed Intra- and interassay variability showed an overall ssues three times. Two-hundred microliters of the super- good reproducibility (coefficient of variation < 5%). I natant were used to extract viral nucleic acids using AdVs were found in the highest concentrations in the NucliSens miniMAG (BioMérieux Italia, Italy), both raw and treated waters (up to 9.8 × 108 GC/ and the RNA was resuspended in a final volume of mL and up to 4.9 × 108 GC/mL, respectively, see 100 μl of elution buffer. Ten microliters of purified Table 2). The viruses found in the lowest concentra- 6

viral nucleic acids were used as template in real-time tions were EVs (2.4 × 10 GC/mL) in inflows, and a l vi ronment PCR amplifications, to measure the increment in NoVs GII (9.9 × 105 GC/mL) in outflows. Mean n GC counts after cell culture, as described above. values (trimmed mean) in raw samples were ranked E The products obtained through ICC-RT-PCR were as follows: AdV > NoV GI > NoV GII > EV, while then analyzed to confirm the presence of EV, and in treated waters, AdV> NoV GI > EV > NoV GII possibly gain additional information regarding the (Figure 1). type of EV found. This was done by running PCR Considering the three groups of viruses together, amplicons (expected size 192 bp) on 2% agarose the total GC count was 4.8 × 107 in influents, and gels containing ethidium bromide (0.5 µg/mL) and 1.2 × 107 in effluents, for a total removal of 74%. visualizing them on a UV transilluminator. Some In an overall analysis of removal efficiency, based representative positive samples were purified and on mean values across months, the viruses most ef- sequenced in a capillary automatic sequencer (ABI ficiently removed were NoV GI (78.4%); while EVs PRISM 310 Genetic Analyzer, Applied Biosystems, were the least efficiently removed (34.6%). Italy) using previously described protocols [16]. Considering sample-pairs, it was noted that some effluent samples were completely cleared of certain viruses (GCs in outflows not detected): 37.5% of ef- RESULTS fluent samples were free of NoV GII, 53.3% were Concentration efficiency free of NoV GI, 8.3% were free of EVs, and 11.1% The efficiency of the method of concentration was were free of AdVs. The majority of effluent samples evaluated by real-time PCR using distilled water were only partially cleared, each to a different ex- spiked with echovirus 2 (in triplicate) and comput- tent. Moreover, in a few, rare cases, outflows were ing the ratio of GC after/before concentration. Out found to be more contaminated than inflows. of 5.4 × 108/reaction of spiked virus, 1.9 × 108 GC/ The viruses most frequently found in samples test- reaction (SD = 7.3 × 107) were recovered, giving a ed by real-time PCR were EVs (96% in influent, and yield of 35.4% ± 3.3%. 84% in effluent samples) and AdVs (96% in influent, and 76% in effluent samples). A smaller proportion Bacteriological results of samples were positive for NoVs GI and GII (92% Quantitative bacteriological data is presented in and 72% at the point of entry and 68% and 40% Table 1. For each microorganism tested, the table at the point of exit, respectively). The simultaneous reports mean, minimum and maximum values for presence of more than one virus was the rule: 64% influents and effluents, as well as percent removal. of influent samples (against 28% of effluent- sam A two-log reduction, or more, in viable bacterial ples) were positive for all tested viruses simultane-

Table 1 | Quantification of viruses and bacterial indicators in wastewater treatment plants

Statistics Total NoV GII NoV GI Enterovirus Adenovirus E. Coli Enterococci IN OUT IN OUT IN OUT IN OUT IN OUT IN OUT IN OUT

Mean 4.80E+07 1.20E+07 2.60E+06 9.90E+05 9.30E+06 2.00E+06 2.40E+06 1.60E+06 3.30E+07 7.60E+06 2.10E+07 2.20E+05 2.9E+06 4.40E+04 Max 5.70E+08 5.60E+07 1.60E+08 1.80E+08 6.80E+06 6.50E+06 9.80E+08 4.90E+08 8.30E+07 1.50E+06 7.70E+06 3.20E+05 Min 6.20E +04 7.80E+02 1.90E+06 1.80E+06 1.10E+05 2.00E+05 4.60E+05 3.40E+05 1.10E+06 5.80E+01 8.40E+05 8.10E+05 Removal -74.00% -62.60% -78.40% -34.60% -77.10% -98.90% -98.50%

Data are expressed in GC/mL for viruses and MPN/100mL for bacteria. Mean is intended as arithmetic mean for bacteria and mean trimmed 20%. for viruses. 270 Giuseppina La Rosa, Manoochehr Pourshaban, Marcello Iaconelli, et al.

Table 2 | Primers and PCRs used in this study

PCR ID Primer ID Sequence (5’ > 3’) Product length (bp) on c ern

C 1080-f CWTACATGCACATCKCSGG 494 1081-r CRCGGGCRAAYTGCACCAG 69 Adenovirus 1082-p fam-CCGGGCTCAGGTACTCCGAGGCGTCCT -tamra e a lt h

H 1064-f GGCCCCTGAATGCGGCTAAT of 377 1065-r CACCGGATGGCCAATCCAA 192 Enterovirus 1066-p fam-CGGACACCCAAAGTAGTCGGTTCCG-tamra ssues

I 1315-f GCCATGTTCCGITGGATG 500 1316-r TCCTTAGACGCCATCATCAT 96 Norovirus GI 1377-p fam-TGTGGACAGGAGATCGCAATCTC -tamra

1380-f CAAGAGYCAATGTTYAGRTGGATGAG 581 1319-r TCGACGCCATCTTCATTCACA 98 Norovirus GII

a l vi ronment 1379-p joe-TGGGAGGGCGATCGCAATCT-tamra n E

ously; 28%, in both raw and treated samples, were of the water environment. Due to their highly infec- positive for three viruses; and 8% of samples were tious nature, viruses are of particular public health positive for 2 viruses at the point of entry (28% at concern. Indeed, they are believed to be the major the point of exit). None of the influent samples were cause of disease contracted through direct contact positive for only one virus (16% in outflows), and with sewage, and are responsible for gastroenteritis, none were completely negative. hepatitis, respiratory illness, and other health prob- lems. Among the wide range of viruses excreted in Enterovirus infectivity and sequence analysis human waste, three groups of enteric viruses – EVs, Nine (out 25) effluent samples showed signs of cy- AdVs, and NoVs – have attracted considerable at- topathic effect on RD cells; GC/liter were calculated tention due to their epidemiological significance as before and after cell culture and the exponential vi- waterborne pathogens [25-27, 13, 8]. ral growth was confirmed by GC counts in cellular The objective of the present study was to assess lysates (at least 4 log). Real-time positive samples the concentration of these viruses in inflows and were confirmed by sequence analysis of the 192-bp outflows, to gain some insight into the infectious amplicon and characterized as EV of the B species potential of viral contamination in treated wastewa- (accession numbers from FN397854 to FN397859). ter, and to obtain a rough estimate of the efficacy of the purification process. Molecular methods of virus quantification were used for this purpose. In DISCUSSION particular, we used TaqMan real-time PCR assays Disposal of inadequately treated wastewater is one which have the advantage of providing rapid, accu- of the main sources of pathogenic contamination rate, sensitive and quantitative detection. For AdVs

�������� �������� �������� �������� �������� �������� �������� �������� �������� �������� ��������

�������� ������������ Fig. 1 | Mean genome copy counts in ������������� ����� �� ������������� ��������������� ���������������� �������������� influent and effluent samples in five �������������� ���������������� ����������������� wastewater treatment plants in and Dark grey = raw sewage; light grey = treated sewage. around Rome. Quantification of enteric viruses at WTPs 271

we used a broad range TaqMan PCR assay previ- GI > EV > NoV GII. Removal efficiencies ranged ously shown to be suitable for quantitative deter- from 35% (EV) to 78% (NoV GII). Considering all mination of viral DNA from all adenovirus Species groups of viruses together, the overall removal ef- (A-F), in both environmental and clinical samples ficiency was 74%. These data are in agreement with on c ern [20]. The presence of AdVs was common in both in- previous studies showing that enteric viruses are C fluent and effluent samples (96%, and 76%, respec- present in high levels not only in raw, but also in tively). Moreover, of all tested viruses, AdVs had the treated wastewaters [4, 5, 39, 7]. e a lt h highest concentrations in both raw and treated wa- Our removal efficiency results are in agreement H of

ters. Overall, the concentration of AdVs was higher with other studies, suggesting that even properly than that of all other viruses combined. This find- working wastewater treatment systems remove on- ing is in agreement with a 2008 study [28], reporting ly about 20-80% of enteric viruses [40-42], viruses ssues

AdV concentrations to be the highest of all enteric which can easily reach water sources and become a I viruses tested, as well as with earlier studies [29, 30]. serious human health concern. The removal of bac- The higher concentrations of AdVs throughout the terial indicators, on the other hand, was much more present study may be due to the fact that AdV, being efficient (up to 98.9%). Our data thus confirm both the only DNA virus, is thus also the most thermosta- the high removal efficiency for bacterial indicators,

ble. Indeed, different studies have shown AdVs to and the absence of correlation between the removal a l vi ronment survive longer than fecal indicator bacteria both in of enteric bacteria and that of enteric viruses, as n sewage and in the environment, and to be very re- widely described in literature [42-44]. E sistant to UV light [31-34]. AdV concentrations re- It should be noted, however, that our samples, col- ported in the present study are slightly higher than lected once a month, were grab, rather than com- those previously reported [35], and may be due to a posite samples (this, by the way, explains the fact higher prevalence of infection in the population. that some samples were more contaminated at the For NoV quantification we used a real-time RT- point of exit). Therefore, our results, rather than be- PCR method previously tested in clinical samples, ing considered representative of the actual removal as well as in estuarine, seawater and sewage water efficiency at the WTP, should instead be regarded samples [18]. NoVs were found to be less prevalent as merely giving a general idea of the amount of vi- than AdVs: 92% and 68% of samples were positive ral contamination at the points of entry and exit. for GII and 72% and 40% for GI in influents and Another limit of the study is in the collection period effluents, respectively. Higher counts for GI as com- (5 months, from May to September), not sufficient pared to GII were detected in both influent and ef- for evaluating seasonal changes in viral concentra- fluent samples. The viral titers obtained in this study tions, especially for noroviruses, responsible for are consistent with previously reported levels [7, 4, gastroenteritis in winter months. However, recent 36], despite the different assay used. studies have shown an increase in spring and sum- The higher concentrations of NoV GI relative to mer peak of NoV outbreaks of gastroenteritis since GII is seemingly at odds with epidemiological data 2002 [37], and an increase of NoV concentration in on the occurrence of NoV outbreaks [37, 38] and WTPs throughout the summer [36]. A detailed anal- sporadic cases [16], mostly attributable to NoV GII. ysis of seasonal changes in viral concentrations was, This finding, suggesting a significant presence of GI however, out of the scope of the present work. in human populations, has already been described by The ability to rapidly detect and quantify viruses in others, both in France [7] and in Sweden [36], and im- environmental samples using real-time PCR repre- plies the need for further studies to address possible sents a considerable advancement holding great po- differences in the pathogenic potential of different ge- tential for environmental applications. Nevertheless, nogroups, their stability in water environments, and the interpretation, in terms of health implications, their ability to evade the human immune system. of the presence of viral genome detected by PCR re- The quantification of EVs was performed using an mains difficult due to the fact that molecular meth- assay based on TaqMan probe hydrolysis technol- ods do not distinguish between infectious and non- ogy in the most conserved region of the enterovi- infectious viral particles. We therefore proceeded to ral genome, which has sufficient genetic variability evaluate the infectivity of viruses in treated effluents to enable identification of the different enteroviral through ICC-RT-PCR. This approach allowed us to types [23]. The proportion of samples positive for overcome most of the disadvantages associated with EVs was the highest (96% and 84% in influents and both conventional cell culture and direct PCR assays, effluents, respectively), but their concentrations were reducing the time needed for the detection of infec- the lowest. This may indicate that EVs are excreted tious viruses. We chose to apply this method to EVs, more consistently, even in cases of asymptomatic because, of the three groups of viruses tested in this infection, while other enteric viruses are mostly as- study, they are the only group that can grow efficient- sociated with outbreaks and therefore excreted less ly on cell lines. The culturing of AdVs (particularly consistently, but at greater concentrations. enteric types 40 and 41) is more problematic, and no To sum up, mean viral genome concentrations were simple cell culture system is available for NoVs at ranked as follows: in influent samples, AdV > NoV present, despite recent efforts [45]. The ICC-real-time GI > NoV GII > EV, and in effluents, AdV > NoV PCR method provided results within 4-5 days of sew- 272 Giuseppina La Rosa, Manoochehr Pourshaban, Marcello Iaconelli, et al.

age-sample collection. Viral growth was confirmed CONCLUSIONS by an exponential increment of GC counts in cellu- This work confirms the frequent occurrence of lar lysates. The results revealed that treated water still enteric viral genomes in sewage effluents, suggesting on c ern contains infectious human EVs (36% of samples) and that treated sewage may represent a source of envi- C thereby represents a potential health hazard. ronmental contamination with potentially infectious As for the other enteric viruses studied, the frequent viruses. Of all tested viruses, AdVs had the highest e a lt h detection of viral genome in sewage effluents suggests concentrations in both raw and treated waters, con-

H that treated sewage may represent a source of envi- firming the relevance of evaluating these viruses as of

ronmental contamination with potentially infectious possible indicators of viral contamination of water. viruses. Indeed, previous studies have detected infec- The real-time PCR has been demonstrated to be tious human AdVs in tertiary-treated and ultravio- a powerful tool for rapid determination of enteric ssues

I let-disinfected wastewaters, confirming the relative viruses in environmental samples and represents a resistance of the AdV isolates [46]. considerable advancement in pathogen quantifica- In recent years, the detection and molecular char- tion in aquatic environments. acterization of pathogenic human viruses in urban sewage have been extensively used to get information Acknowledgments

a l vi ronment on circulating viruses in given populations through- n out the world; however to the best of our knowledge, This study was partly funded by the project “Rapid diagnosis of E no previous investigations have specifically addressed viruses in sewage” (CCM n. 4393, ISS Prot. 8M09) of the Italian Center for Disease Control and Prevention – Centro Nazionale and compared viral concentrations in WTPs for dif- per la Prevenzione e il Controllo delle Malattie (CCM) – and the ferent epidemiologically important, waterborne en- Italian Ministry of Health. We are also particularly grateful to teric viruses, using quantitative assays. Simona Sermoneta, Master of Public Health, for the linguistic Given the considerable burden of viral gastrointes- and critical revision of the manuscript. tinal disease, and in the absence of costly epidemio- logical monitoring systems, virological monitoring of sewage flowing into a WTP may represent an inexpen- Conflict of interest statement sive and efficient approach that would allow for an There are no potential conflicts of interest or any financial or per- ongoing follow-up on the occurrence of pathogenic sonal relationships with other people or organizations that could inappropriately bias conduct and findings of this study. viruses in the plant’s service area. The monitoring of effluent samples, on the other hand, may contribute to the understanding of viral removal through waste- Submitted on invitation. water treatment and to the prevention of disease. Accepted on 22 April 2010.

References 1. Bosch A. Human enteric viruses in the water environment: a lance through monitoring of sewer systems. Adv Appl Microbiol minireview. Int Microbiol 1998;1:191-6. 2008;65:249-69. 2. Hejkal TW, Wellings FM, Lewis AL, LaRock PA. distribu- 10. Pasquinelli L. Enterovirus infections. Pediatr Rev 2006;27: tion of viruses associated with particles in waste water. Appl e14-e15. Environ Microbiol 1981;41:628-34. 11. Palacios G, Oberste MS. Enteroviruses as agents of emerging 3. Blatchley ER, III, Gong WL, Alleman JE, Rose JB, Huffman infectious diseases. J Neurovirol 2005;11:424-33. DE, Otaki M, et al. Effects of wastewater disinfection on water- 12. Lenaerts L, De CE, Naesens L. Clinical features and treat- borne bacteria and viruses. Water Environ Res 2007;79: 81-92. ment of adenovirus infections. Rev Med Virol 2008;18:357-74. 4. Laverick MA, Wyn-Jones AP, Carter MJ. Quantitative RT- PCR for the enumeration of noroviruses (Norwalk-like virus- 13. La Rosa G, Pourshaban M, Iaconelli M, Muscillo M. es) in water and sewage. Lett Appl Microbiol 2004;39:127-36. Recreational and drinking waters as a source of norovi- rus gastroenteritis outbreaks: a review and update. Environ 5. van den Berg H, Lodder W, van der Poel W, Vennema H, de Biotechnol 2008;4:15-24. Roda Husman AM. Genetic diversity of noroviruses in raw and treated sewage water. Res Microbiol 2005;156:532-40. 14. Patel MM, Hall AJ, Vinje J, Parashar UD. Noroviruses: a comprehensive review. J Clin Virol 2009;44:1-8. 6. Haramoto E, Katayama H, Oguma K, Koibuchi Y, Furumai H, Ohgaki S. Effects of rainfall on the occurrence of human 15. Zheng DP, Ando T, Fankhauser RL, Beard RS, Glass RI, adenoviruses, total coliforms, and Escherichia coli in seawa- Monroe SS. Norovirus classification and proposed strain no- ter. Water Sci Technol 2006;54:225-30. menclature. Virology 2006;346:312-23. 7. da Silva AK, Le Saux JC, Parnaudeau S, Pommepuy M, 16. La Rosa G, Fontana S, Di Grazia A, Iaconelli M, Pourshaban Elimelech M, Le Guyader FS. Evaluation of removal of M, Muscillo M. Molecular Identification and genetic analy- noroviruses during wastewater treatment, using real-time re- sis of norovirus genogroups I and II in water environments: verse transcription-PCR: different behaviors of genogroups comparative analysis of different reverse transcription-PCR I and II. Appl Environ Microbiol 2007;73:7891-7. Assays. Appl Environ Microbiol 2007;73:4152-61 [Erratum in: Appl Environ Microbiol 2007;73(19):6329]. 8. Fong TT, Lipp EK. Enteric viruses of humans and animals in aquatic environments: health risks, detection, and poten- 17. Boom R, Sol CJ, Salimans MM, Jansen CL, Wertheim-van tial water quality assessment tools. Microbiol Mol Biol Rev Dillen PM, van der NJ. Rapid and simple method for purifi- 2005;69:357-71. cation of nucleic acids. J Clin Microbiol 1990;28:495-503. 9. Sinclair RG, Choi CY, Riley MR, Gerba CP. Pathogen surveil- 18. La Rosa G, Pourshaban M, Iaconelli M, Muscillo M. Quantification of enteric viruses at WTPs 273

Quantification of Norovirus genogroups I and II in envi- 33. Thurston-Enriquez JA, Haas CN, Jacangelo J, Gerba CP. ronmental and clinical samples using Taqman real-time RT- Inactivation of enteric adenovirus and feline calicivirus by PCR. Food Environ Virol 2009;1:15-22. ozone. Water Res 2005;39:3650-6.

19. Jothikumar N, Lowther JA, Henshilwood K, Lees DN, Hill 34. Thurston-Enriquez JA, Haas CN, Jacangelo J, Gerba CP. on c ern

VR, Vinje J. Rapid and Sensitive detection of noroviruses by Inactivation of enteric adenovirus and feline calicivirus by C using TaqMan-based one-step reverse transcription-PCR as- chlorine dioxide. Appl Environ Microbiol 2005;71:3100-5. says and application to naturally contaminated shellfish sam- 35. Bofill-Mas S, Albinana-Gimenez N, Clemente-Casares P, ples. Appl Environ Microbiol 2005;71:1870-5. Hundesa A, Rodriguez-Manzano J, Allard A, et al. wuan- e a lt h 20. Muscillo M, Pourshaban M, Iaconelli M, Fontana S, Di tification and stability of human adenoviruses and polyoma- H of

Grazia A, Manzara S, et al. Detection and quantification virus JCPyV in wastewater matrices. Appl Environ Microbiol of human adenoviruses in surface waters by Nested PCR, 2006;72:7894-6. TaqMan real-time PCR and cell culture assays. Wat Air Soil 36. Nordgren J, Matussek A, Mattsson A, Svensson A, Lindgren Pollut 2008;191:1-4. ssues P. Prevalence of norovirus and factors influencing virus con- I 21. Hernroth BE, Conden-Hansson AC, Rehnstam-Holm AS, centrations during one-year in a fullscale wastewater treat- Girones R, Allard AK. Environmental factors influencing ment plant. Water Res 2009;43:1117-25. human viral pathogens and their potential indicator organ- 37. Lopman B, Vennema H, Kohli E, Pothier P, Sanchez A, isms in the blue mussel, Mytilus edulis: the first Scandinavian Negredo A, et al. Increase in viral gastroenteritis outbreaks report. Appl Environ Microbiol 2002;68:4523-33. in Europe and epidemic spread of new norovirus variant.

22. Muscillo M, Aulicino FA, Patti AM, Orsini P, Volterra L, Lancet 2004;363:682-8. a l vi ronment Fara GM. Molecular techniques for the identification of en- n 38. Lodder WJ, Roda Husman AM. Presence of noroviruses E teric viruses in marine waters. Wat Res 1994;28:1-7. and other enteric viruses in sewage and surface waters in The 23. Donaldson KA, Griffin DW, Paul JH. Detection, quantita- Netherlands. Appl Environ Microbiol 2005;71:1453-61. tion and identification of enteroviruses from surface waters 39. Haramoto E, Katayama H, Oguma K, Yamashita H, Tajima and sponge tissue from the Florida Keys using real-time RT- A, Nakajima H, et al. Seasonal profiles of human norovi- PCR. Water Res 2002;36(10):2505 -14. ruses and indicator bacteria in a wastewater treatment plant 24. Yuan JS, Reed A, Chen F, Stewart CN Jr. Statistical analysis in Tokyo, Japan. Water Sci Technol 2006;54:301-8. of real-time PCR data. BMC Bioinformatics 2006;7:85. 40. Payment P, Fortin S, Trudel M. Elimination of human en- 25. Sinclair RG, Jones EL, Gerba CP. Viruses in recreational wa- teric viruses during conventional waste water treatment by ter-borne disease outbreaks: a review. J Appl Microbiol 2009; activated sludge. Can J Microbiol 1986;32:922-5. 107:1769-80. 41. Powelson DK, Gerba CP. Viral removal from sewage efflu- 26. Maunula L (Ed.).Waterborne norovirus outbreak. Future Virol ents during saturated and unsaturated flow through soil col- 2007;2:101-12. umns. Water Res 1994;28:2175-81. 42. Ottoson J, Hansen A, Bjorlenius B, Norder H, Stenstrom 27. Mena KD, Gerba CP. Waterborne adenovirus. Rev Environ TA. Removal of viruses, parasitic protozoa and microbial in- Contam Toxicol 2009;198:133-67. dicators in conventional and membrane processes in a waste- 28. Katayama H, Haramoto E, Oguma K, Yamashita H, Tajima water pilot plant. Water Res 2006;40:1449-57. A, Nakajima H, et al. One-year monthly quantitative survey 43. Fleischer J, Schlafmann K, Otchwemah R, Botzenhart K. of noroviruses, enteroviruses, and adenoviruses in wastewater Elimination of enteroviruses, other enteric viruses, F-specific collected from six plants in Japan. Water Res 2008;42:1441-8. coliphages, somatic coliphages and E. coli in four sewage 29. Pina S, Puig M, Lucena F, Jofre J, Girones R. Viral pollution treatment plants of Southern Germany. Water Supply Res in the environment and in shellfish: human adenovirus de- Technol 2000;49:127-37. tection by PCR as an index of human viruses. Appl Environ 44. Jacangelo JG, Loughran P, Petrik B, Simpson D, McIlroy C. Microbiol 1998;64:3376-82. Removal of enteric viruses and selected microbial indicators 30. Irving NLG, Smith FA. One-year survey of enteroviruses, by UV irradiation of secondary effluent. Water Sci Technol adenoviruses, and reoviruses isolated from effluent at an 2003;47:193-8. activated-sludge purification plant. Appl Environ Microbiol 45. Straub TM, Honer zu BK, Orosz-Coghlan P, Dohnalkova 1981;41:51-9. A, Mayer BK, Bartholomew RA, et al. In vitro cell culture 31. Enriquez CE, Hurst CJ, Gerba CP. Survival of the enteric infectivity assay for human noroviruses. Emerg Infect Dis adenoviruses 40 and 41 in tap, sea, and waste water. Water 2007;13:396-403. Res 1995;29:2548-53. 46. Thompson SS, Suva-Castillo M, Yanko WA, El Jack Z, Kuo 32. Gerba CP, Gramos DM, Nwachuku N. Comparative inacti- J, Chen CL, et al. Detection of infectious human adenovirus- vation of enteroviruses and adenovirus 2 by UV light. Appl es in tertiary-treated and ultraviolet-disinfected wastewater. Environ Microbiol 2002;68:5167-9. Water Environ Res 2003;75:163-70.